Quenching of singlet oxygen by oxygen- and sulfur-centered radicals: Evidence for energy transfer to peroxyl radicals in solution

Article abstract of DOI:10.1021/ja9730831

Quenching of singlet oxygen luminescence at 1.27 μm by PhS^., PhSO^., and peroxyl radicals PhOO^., t-BuOO^., PhCH₂OO^., Ph₂CHOO^. and Ph₃COO^. was studied in liquid solution. The quantum yields of decomposition of different initiators which lead to the formation of free radicals were measured by using nanosecond transient absorption. This allowed determination of singlet oxygen O₂(¹Δ(g)) quenching rate constants by the radicals. They are <2 x 10⁸ M^-1 s^-1 for the sulfur-centered radicals and (2-7) x 10⁹ M^-1 s^-1 for peroxyl radicals in acetonitrile. The rapid quenching is attributed to energy transfer quenching by the peroxyls, which have an n → π* transition leading to a low-lying ²A' state above their ²A'' ground state. PhSO^. is shown computationally not to have such a low-lying ²A' state. There may be a very low-lying ²B₁ state, for PhS^., but it is apparently not an efficient acceptor of electronic energy from O₂(¹Δ(g)).

Full text of DOI:10.1021/ja9730831

396
J. Am. Chem. Soc. 1998, 120, 396-403
Quenching of Singlet Oxygen by Oxygen- and Sulfur-Centered
Radicals: Evidence for Energy Transfer to Peroxyl Radicals in
Solution
Alexandre P. Darmanyan,† Daniel D. Gregory,† Yushen Guo,† William S. Jenks,*,†
Laure Burel,‡ Dominique Eloy,‡ and Pierre Jardon‡
Contribution from the Department of Chemistry, Iowa State UniVersity, Ames, Iowa 50011-3111, and
Laboratoire D’e´lectrochimie Organique et de Photochimie Redox, UniVersite of Joseph Fourier,
B.P. 53-38041 Grenoble, Cedex, France
ReceiVed September 2, 1997X
Abstract: Quenching of singlet oxygen luminescence at 1.27 µm by PhS•, PhSO•, and peroxyl radicals PhOO•,
t-BuOO•, PhCH2OO•, Ph2CHOO•, and Ph3COO• was studied in liquid solution. The quantum yields of
decomposition of different initiators which lead to the formation of free radicals were measured by using
nanosecond transient absorption. This allowed determination of singlet oxygen O2(1∆g) quenching rate constants
by the radicals. They are <2 × 108 M-1 s-1 for the sulfur-centered radicals and (2-7) × 109 M-1 s-1 for
peroxyl radicals in acetonitrile. The rapid quenching is attributed to energy transfer quenching by the peroxyls,
2
2
which have an n f π* transition leading to a low-lying A′ state above their A′′ ground state. PhSO• is
2
2
shown computationally not to have such a low-lying A′ state. There may be a very low-lying B1 state, for
PhS•, but it is apparently not an efficient acceptor of electronic energy from O2(1∆g).
Introduction
spectra for HOO•, CH3OO•, C2H5OO•, and (CH3)2CHOO• have
been recorded,10,11 and the excitation energies all lie below that
Peroxyl radicals (ROO•) are the principal intermediates and
chain propagators in autoxidation reactions. They oxidize
organic and biological compounds by mechanisms of electron
transfer and hydrogen abstraction.1,2 Singlet oxygen (O2(1∆g))
is another general photooxidation intermediate, and it is also a
very powerful oxidant of polymers.3-5 In polybutadiene and
polyisoprene, for instance, the lifetimes of O2(1∆g) are 3.6 and
2.6 µs, respectively, and hydroperoxides are formed as a result
of chemical quenching.6,7 Both singlet oxygen and organic
radicals are generally cytotoxic, and the former has been put to
use in the form of photodynamic therapy of tumors. Thus it is
important to know how singlet oxygen interacts with different
free radicals in the study of photooxidation processes.
of singlet oxygen, which is 7884 cm-1 4
.
Additionally, quench-
ing of O2(1∆g) luminescence in benzene by acylperoxyl radicals
•
•
•
CH3COO2 , PhCOO2 , and p-CH3C6H4COO2 with rate constants
all nearly 1 × 1010 M-1 s-1 has been reported.12
These observations led us to believe that the peroxyl radical
might be an efficient quencher of singlet oxygen regardless of
substituent and to pursue the experiments described herein.
Additionally, our recent interest in sulfoxide photochemistry13
has brought us experience with sulfinyl radicals (RSO•),14 whose
relationship to peroxyl radicals is obvious. Much less is known
about these species in general, but no excited state lying below
8000 cm-1 is known for HSO•, which implied that there might
be decidedly contrasting behavior between the peroxyl and
sulfinyl species toward singlet oxygen.
The quenching of O2(1∆g) by HOO• has been observed in
the gas phase by Becker et al., and it was suggested that the
mechanism was energy transfer.8 The ground state of simple
peroxyl radicals is 2A′′ , and the first excited state (2A′) of HOO•
Nitroxyl radicals do not have low-lying doublet or quartet
excited states, but their stability recommends them for kinetic
investigations. They quench O2(1∆2) in solution with rate
constants far below the diffusion controlled limit: ∼105 - 107
is only 7041 cm-1 higher.9,10 In fact, A′′ f A′ absorption
2
2
† Iowa State University.
‡ Universite of Joseph Fourier.
M-1 s-1 15
. It was concluded that the quenching was by
X Abstract published in AdVance ACS Abstracts, December 15, 1997.
(1) Semenov, N. N. Some Problems of Chemical Kinetics and ReactiVity;
Pergammon Press: New York, 1958; Vols. 1 and 2.
(2) Ingold, K. U. Acc. Chem. Res 1969, 2, 1-9.
(3) Wilkinson, F.; Helman, W. P.; Ross, A. B. J. Phys. Chem. Ref. Data
1995, 24, 663-1020.
1
-
enhancement of the
∆ f
g
3Σg transition by an electron
exchange mechanism. This stands as a “baseline” mechanism
(10) Hunziker, H. E.; Wendt, H. R. J. Chem. Phys. 1976, 64, 3488-
3490.
(11) Jacox, M. E. J. Phys. Chem. Ref. Data 1994, Monograph 3, 1-461.
(12) Darmanyan, A. P.; Foote, C. S.; Jardon, P. J. Phys. Chem. 1995,
99, 11854-11859.
(4) Frimer, A. A. Singlet O2; CRC Press: Boca Raton, FL, 1985.
(5) Ranby, B.; Rabek, J. F. Photodegradation, Photooxidation, and
Photostabilization of Polymers:
Principles and Application;
Wiley-Interscience: London, 1975.
(13) Jenks, W. S.; Gregory, D. D.; Guo, Y.; Lee, W.; Tetzlaff, T. In
Organic Photochemistry; Ramamurthy, V., Schanze, K. S., Eds.; Marcel
Dekker, Inc.: New York, 1997; Vol. 1, pp 1-56.
(14) Darmanyan, A. P.; Gregory, D. D.; Guo, Y.; Jenks, W. S. J. Phys.
Chem. A 1997, 101, 6855-6863.
(6) Matveev, Y. Visokomol. Soedin, Ser. B 1990, 32, 441-444.
(7) Darmanyan, A. P.; Matveev, M. Y. SoV. J. Chem. Phys. 1990, 5,
2768-2786.
(8) Becker, K. H.; Fink, E. H.; Langen, P.; Scurath, U. J. Chem. Phys.
1974, 60, 4623-4625.
(15) Darmanyan, A. P.; Tatikolov, A. S. J. Photochem. 1986, 32, 157-
163.
(9) Hunziker, H. E.; Wendt, H. R. J. Chem. Phys. 1974, 60, 4622-4623.
S0002-7863(97)03083-7 CCC: $15.00 © 1998 American Chemical Society
Published on Web 01/21/1998

Quenching of Singlet O2 by O- and S-Centered Radicals
J. Am. Chem. Soc., Vol. 120, No. 2, 1998 397
for singlet oxygen quenching that ought to be available for all
radicals in the absence of other, faster interactions.
Here we report the quenching rate constants of O2(1∆g) by
sulfenyl (thiyl), sulfinyl, and peroxyl radicals in solution. We
suggest that the peroxyl radical function indeed is an effective
energy transfer quencher of singlet oxygen. The sulfur centered
radicals quench singlet oxygen much more slowly than the
peroxyl radicals.
Chart 1
Experimental Section
The O2(1∆g) luminescence at 1.27 µm was recorded with a high-
sensitivity Ge detector described elsewhere.15 The fourth harmonic of
Nd:YAG laser Surelite-I (266 nm, 5 ns, 4 mJ, diameter of laser pulse
0.6 mm) was used for the excitation of solutions. Spectra and decay
kinetics of free radicals were recorded by nanosecond laser photolysis.
The yield of alkyl and sulfur-centered radicals from photolysis of the
initiators was measured in Ar-flushed solutions at the absorption
maximum of radicals by comparison to the triplet-triplet absorption
of anthracene in cyclohexane, as previously reported.14 The yield of
free radicals and the intensity of triplet-triplet absorption of anthracene
were linear for laser energies of 1-8 mJ. The optical densities of
solutions were 0.2-0.3 at 266 nm. The solutions in a quartz cell (1 ×
1 cm) were changed after 1 or 2 laser pulses. The energy of the laser
pulse was measured by a Ophir DGX-RP power meter. The accuracy
of quantum yields is estimated to be (20%.
Absorption spectra were recorded on a UV-2101 PC Shimadzu
spectrophotometer. The solutions were deaerated by Ar bubbling for
20 min. Experiments were carried out at ambient temperature, 25 °C.
Spectro grade solvents were used as received. Dibenzyl ketone (1),
chlorotriphenylmethane (2), 5-chlorodibenzosuberane (3), diphenyl
sulfoxide (8), diphenyl disulfide (10), 5,10,15,20-tetraphenylporphine
(TPP), anthracene (AN), and acridine (AC) were obtained commercially
and purified as necessary by recrystallization. Phenyl diphenylmethyl
sulfoxide (4), phenyl diphenylmethyl sulfide (5), p-tolyl triphenylmethyl
sulfide (6), phenyl benzenethiosulfinate (7), and tert-butyl phenyl
sulfoxide (9) were prepared by literature methods.17-21
Results
Rate Constants. Photolysis of dibenzyl ketone (1), chloro-
triphenylmethane (2), and 5-chlorodibenzosuberane (3) leads
to the production of benzyl, triphenylmethyl, and 5-dibenzo-
suberanyl (3•) alkyl radicals with well-established absorption
spectra and extinction coefficients.25-30 The quantum yields
of decomposition of these initiators leading to the production
of free radicals (Φesc) were unpublished, except for 1, and were
determined by using the published extinction coefficients (Table
1).26,29,31
The excitation of dibenzyl ketone leads to very rapid
production of two PhCH2 radicals as a result of the initial
•
R-cleavage and rapid decarbonylation of the phenylacetyl
radical:32-34
PhCH2COCH2Ph + hν f PhCH2CO• + PhCH2 (1)
•
Computations were carried out with the GAMESS suite of pro-
grams.22 All radicals were fully optimized at the ROHF/6-31G(d,p)
level and were found to be true minima, except for PhSO•, where
transition states and minima were found, as noted in the text. For
PhSO•, CASSCF single point energies were determined at the ROHF
geometries using active spaces corresponding to the π structure on the
phenyl ring plus two “lone pair” orbitals on S and three on O (i.e., 15
electrons in 8 a′′ orbitals and 3 a′ orbitals for the planar structure and
7 a′ and 4 a′′ orbitals for the T-structure; see the text for descriptions
of these structures). These showed that the doubly occupied S and O
lone pairs were not appropriate for inclusion in the CASSCF wave
function, and the calculations were re-run including only the phenyl
π-system and the singly occupied S-O π* orbital (2A′′ state) or
O-centered p-orbital (2A′ state). These were 7 electrons in 7 orbitals.
For the planar A′ state, the singly occupied a′ orbital also had to be
included (7 electrons in 8 orbitals), but it has an occupation of 1.000
due to symmetry. Quasidegenerate multireference second-order per-
turbation theory (MRMP2)23,24 corrections were then used to recover
the dynamic correlation energies for the multireference wave functions.
PhCH2CO• f PhCH2 + CO
(2)
•
In acetonitrile, Φesc was found to be 0.85 ( 0.17. This estimate
is within reasonable agreement of the previously reported value
of 0.71 ( 0.11.25
Ph2CH•, Ph3C•, and 3• undergo solvent independent photo-
chemical reactions from the excited doublet states.28,29 How-
ever, these radicals have low extinction coefficients at 266 nm,
and thus two-photon decomposition and photochemical reactions
of arylmethyl radicals from the excited states were neglected.
•
In contrast, PhCH2 strongly absorbs at 266 nm (ꢀ ∼ 8 × 103
M-1 cm-1 26
and the generation and decay of this radical in
)
cyclohexane and CH3CN was investigated. In agreement with
previous reports, we did not find any decomposition products
(25) Robbins, W. K.; Eastman, R. H. J. Am. Chem. Soc. 1970, 92, 6076-
6077.
(16) Mitsui, M.; Takeda, K.; Kobori, Y.; Kawai, A.; Obi, K. Chem. Phys.
Lett. 1996, 262, 125-130.
(26) Huggenberger, C.; Fischer, H. HelV. Chim. Acta 1981, 338-353.
(27) Claridge, R. F. C.; Fischer, H. J. Phys. Chem. 1983, 87, 1960-
(17) Backer, H. J.; Kloosterziel, H. Recl. TraV. Chim. 1954, 73, 129-
239.
1967.
(28) Bromberg, A.; Meisel, D. J. Phys. Chem. 1985, 89, 2507-2513.
(29) Bromberg, A.; Schmidt, K. H.; Meisel, D. J. Am. Chem. Soc. 1985,
(18) Balicki, R.; Kaczmarek, L.; Nantka-Namirski, P. Liebigs Ann. Chem.
1992, 883-884.
(19) Finzi, C.; Bellavita, V. Gazz. Chim. Ital. 1933, 62, 699-708.
(20) Guo, Y.; Jenks, W. S. J. Org. Chem. 1997, 62, 857-864.
(21) Kozuka, S.; Takahashi, H.; Oae, S. Bull. Chem. Soc. Jpn. 1970, 43,
129-134.
(22) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.;
Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, N.; Su,
S. J.; Windus, T. L.; Dupuis, M.; Montgomery, J. A. J. Comput. Chem.
1993, 14, 1347-1363.
(23) Nakano, H. J. Chem. Phys. 1993, 99, 7983-7992.
(24) Nakano, H. Chem. Phys. Lett. 1993, 207, 372-378.
107, 83-91.
(30) Scaiano, J. C.; Tanner, M.; Weir, D. J. Am. Chem. Soc. 1985, 107,
4396-4403.
(31) Taub, I. A.; Harter, D. A.; Sauer, M. C., Jr.; Dorfman, L. M. J.
Chem. Phys. 1964, 41, 979-985.
(32) Lunazzi, L.; Ingold, K. U.; Scaiano, J. C. J. Phys. Chem. 1983, 87,
529-530.
(33) Turro, N. J.; Gould, I. R.; Baretz, B. H. J. Phys. Chem. 1983, 87,
531-532.
(34) Arbour, C.; Atkinson, G. H. Chem. Phys. Lett. 1989, 159, 520-
525.

398 J. Am. Chem. Soc., Vol. 120, No. 2, 1998
Darmanyan et al.
Table 1. Rate Constants of O2(1∆g) Quenching by Free Radicals and Photophysical Properties of Initiators
a
kinetically important
radical(s)
ꢀ(R•, λmax),
O2(1∆g)
kq,
entry no.
initiator
solvent
hexane
cyclohexane
cyclohexane
hexane
CH3CN
CH3CN
CH3CN
CH3CN
104 M-1 cm-1
Φesc
0.65b,c
0.2
sensitizer
109 M-1s-1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
PhS•
10
7
1
0.9 (PhS•, 295 nm)c
1.1 (PhSO•, 300 nm)c
TPP
AN
AN
TPP
AN
ANe
AN
AN
AC
AN
AN
AN
AC
AN
AC
<0.1
<0.2
4.2
6.3
3.4
5.3
6.5
3.3
4.1
2.6
2.1
7.0
5.2
6.2
7.1
PhSO• + PhS•
•
PhCH2OO•
0.88 (PhCH2 , 316 nm)d
0.85
PhOO•
8
9
4
0.06b,c
0.12b,c
0.18c
t-BuOO•
Ph2CHOO•
4.4 (Ph2CH•, 331 nm)f
5
6
2
CH3CN
CH3CN
CH3CN
0.30
0.38
0.17
Ph3COO•
3.6 (Ph3C•, 338 nm)g
3.6 (Ph3C•, 338 nm)g
Ph3COO• + NCCH2OO•
3-OO•+ NCCH2OO•
3
CH3CN
2.2 (3•, 333 nm)f
0.24
a (30%. b In cyclohehane. c Reference 14. d Reference 26. e In O2-saturated solvent. f Reference 29. g Reference 31.
•
of benzyl radical.35 The decay traces for PhCH2 are well fit
to second-order decay kinetics; the recombination rate constant
estimated at 258 nm is 4.8 × 109 M-1 s-1 in cyclohexane,
similar to previous values.27
The photolysis of sulfur-containing compounds is another
source of free radicals.13,14 Photolysis of diphenyl disulfide
leads to two molecules of PhS•. Similarly, excitation of sulfides
5 and 6 leads to the production of sulfenyl (thiyl) radicals and
•
•
carbon-centered radicals CHPh2 and CPh3 with absorption
maxima at 331 and 338 nm, respectively.29,31
Photolysis of thiosulfinate 7 generates pairs of phenylsulfenyl
and phenylsulfinyl radicals and photolysis of sulfoxide 4 yields
similar R-cleavage products:
PhS(O)-SPh + hν f PhSO• + PhS•
(3)
(4)
PhS(O)-CHPh2 + hν f PhSO• + Ph2CH•
Figure 1. Dependence of O2(1∆g) decay rate constant on concentration
of free radicals: (O) PhCH2OO• in hexane, In is 1, Sens ) TPP; (×)
PhCH2OO• in CH3CN, In is 1, Sens)AN; (4) PhS• in hexane, In is
10, Sens ) TPP.
The value of Φesc was previously estimated for 4.14 Photolysis
of Ph2SO (8) and tert-butyl phenyl sulfoxide (9) occurs
analogously, yielding PhSO• with Ph• or (CH3)3C•, respec-
tively.14 All Φesc results are reported in Table 1.
Carbon-centered radicals react very quickly by addition with
molecular oxygen to form peroxyl radicals (eq 5).36-38
The one exception is phenyl, due to its higher reactivity with
solvent. The rate constant for hydrogen atom abstraction from
CH3CN by phenyl is 1.0 × 105 M-1 s-1 40
.
The reactivity of
kox
9
R• + O2
8 ROO•
(5)
phenyl with O2 (kox), however, is not much different from that
of the benzylic radicals: kox ) 4.4 × 109 M-1 s-1 37
.
Because
The rate constants of reaction 5 for Ph3C• and the dibenzosub-
eranyl radical 3• were unknown and were qualitatively estimated
here. In air-saturated acetonitrile the lifetimes of Ph3C• and 3•
are 900 and 520 ns, respectively. Since the concentration of
oxygen in air-saturated CH3CN is 1.9 × 10-3 M,39 kox is
estimated at 5.8 × 108 and 1.0 × 109 M-1 s-1 for Ph3C• and 3•,
respectively. Given these and the previously known values for
kox,36-38 carbon-centered radicals are converted to peroxyl
radicals in air-saturated solutions within 1 µs of the laser pulse.
of this, to ensure that at least 95% of the phenyl radicals are
converted to PhOO•, O2-saturated solutions had to be used. The
source of phenyl radicals was diphenyl sulfoxide (8).
The sulfur-centered radicals PhS• and PhSO• undergo reaction
5 with rate constants <107 M-1 s-1 14
Thus, even under
.
conditions of air or O2 saturation, neither forms significant
peroxyl-type species within the singlet oxygen lifetime, and
therefore the sulfenyl and sulfinyl radicals are assumed to remain
intact for purposes of the current experiments.
(35) Meisel, D.; Das, P. K.; Hug, G. L.; Bhattacharyya, K.; Fessenden,
R. W. J. Am. Chem. Soc. 1986, 108, 4706-4710.
Singlet oxygen was generated by excitation of sensitizers TPP,
AN, and AC in air-saturated solutions. The lifetimes of O2(1∆g)
in the investigated solvents agreed well with literature data.4
The addition of initiators 1-9 to these solutions led to the sharp
decrease of O2(1∆g) lifetime (Figures 1-3) due to the simul-
taneous formation of radicals which serve as quenchers of 1O2.
(36) Howard, J. A.; Scaiano, J. C. In Landolt-Bo¨rnstein. Numerical Data
and Functional Relationships in Science and Technology. New Series;
Fischer, H., Ed.; Springer-Verlag: Berlin, 1984; Vol. 13d, p 431.
(37) Lusztyk, J.; Kanabus-Kaminska, J. M. In Handbook of Organic
Photochemistry; Scaiano, J. C., Ed.; CRC Press: Boca Raton, FL, 1989;
Vol. 2, pp 177-209.
(38) Maillard, B.; Ingold, K. U.; Scaiano, J. C. J. Am. Chem. Soc. 1983,
105, 5095-5099.
(39) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photochem-
istry, 2nd ed.; Marcel Dekker, Inc.: New York, 1993; p 420.
(40) Scaiano, J. C.; Stewart, L. C. J. Am. Chem. Soc. 1983, 105, 3609-
3614.

Quenching of Singlet O2 by O- and S-Centered Radicals
J. Am. Chem. Soc., Vol. 120, No. 2, 1998 399
The concentration of peroxyl radicals produced by the laser
pulse was 1-10 µM in most cases, and about 30 µM maximum.
Under these conditions, peroxyl radicals decay in solution
predominantly by self-recombination, neglecting momentarily
any interactions with singlet oxygen. The highest known value
for such recombination is that of benzyl peroxyl radicals in
toluene and is about 3 × 108 M-1 s-1 36
.
Taking this value as
an upper limit, the minimal half-life of peroxyl radicals can be
estimated at g300 µs (for 10 µM ROO•) in these experiments.
This is almost an order of magnitude longer than the lifetime
of O2(1∆g) in the absence of the radical quenchers (Figures 1-3).
It is possible that for initiators 4-9 that the peroxyl radicals
have another significant decay channel by reaction with the
sulfinyl and sulfenyl radicals, but this does not change the
conclusion that the lifetime of the peroxyl radicals is much
longer than the lifetime of the singlet oxygen under all
conditions used. Additionally, it was found experimentally that
the 1O2 emission traces were well-fit to first-order decays. Thus
an assumption can be made that the concentration of peroxyl
radicals is constant throughout the decay of the singlet oxygen
emission. Equation 7 was thus used directly to estimate the
concentration of radicals for each datum used to obtain the
quenching rate constants.
Figure 2. Dependence of O2(1∆g) decay rate constant on concentration
of free radicals: (O) (CH3)3COO• in CH3CN, In is 9, Sens ) AN; (×)
Ph2CHOO• in CH3CN, In is 4, Sens ) AN; (4) PhSO• and PhS• in
cyclohexane, In is 7, Sens ) AN.
The sulfinyl and sulfenyl radicals undergo self-quenching
significantly faster than do the peroxyl radicals.14 Because of
this, the first half-life of PhS• and PhSO• is only longer than
the lifetime of singlet oxygen at initial concentrations of up to
a few µM. Despite this, initial concentrations (determined from
eq 7) of up to 30 µM PhS• were generated from PhSSPh. No
change in the singlet oxygen decay was detected. Thus, while
no rate constant could be determined, upper limits for the rate
constant kq were obtained (vide infra).
In principle, singlet oxygen can interact with the initiators in
the ground state as well as the radicals produced by their
photolysis. To determine if these processes were kinetically
important, additional control experiments were carried out.
Singlet oxygen lifetimes were measured by excitation of the
sensitizers at 355 nm, where none of the initiators absorb. The
addition of any of the initiators in concentrations up to 500 µM
did not decrease the lifetime of O2(1∆g), implying that all of
the initiators react with singlet oxygen with rate constants <107
Figure 3. Dependence of O2(1∆g) decay rate constant on concentration
of peroxyl radicals in acetonitrile: (O) 3-OO•, In is 3, Sens ) AN;
(×) Ph3COO• radical, In is 2, Sens ) AC; (4) Ph3COO• radical, In is
6, Sens ) AN.
M-1 s-1
. Therefore, because all further experiments were
carried out with initiator concentrations <500 µM, possible
quenching contributions to kq from this source were ignored.
The concentration of radicals formed as a result of photolysis
of the initiators in the presence of the sensitizer can be
determined from the laser energy and the optical density of each
species. The energy of the laser pulse absorbed by initiator
was estimated from the expression
The O2(1∆g) lifetimes were measured beginning 5 µs after
the laser pulse to avoid the kinetic influences of the following
processes: quenching of the triplet sensitizer by O2, quenching
of O2(1∆g) by the triplet sensitizer,41,42 and reaction 5. Finally,
it had to be assumed that radical concentrations were homoge-
neous across the irradiated volume, which is of necessity an
approximation because of the spatial distribution of the beam
and some absorption of the light as it passes through the cell.
Nevertheless, plots of the inverse of the O2(1∆g) lifetimes versus
the estimated radical concentrations were quite linear and kq
values were obtained for each system. The kq values obtained
by this method and using these approximations are presented
in Table 1.
266(In+Sen)
E1‚OD266(In)[1 - 10OD
]
Eabs
)
(6)
OD266(In+Sen)
where El ) the energy of the laser pulse (4 mJ) and OD266
-
(In+Sen) is less than 1 and is the optical density of the solution
with the initiator added.
The concentration of radicals created after laser pulse was
estimated from eq 7, where V ) 0.28 cm3 is the volume of
irradiated area and ΦR is the quantum yield of radical pairs
escaping geminate recombination.
Estimation of Excited-State Energies for Radicals. It is
evident from Table 1 that the kq values for the peroxyl radicals
are qualitatively similar and at least an order of magnitude
[R•]o ) EabsΦRV-1
(7)
(41) Duncan, C. K.; Kearnes, D. R. Chem. Phys. Lett. 1971, 12, 306-
309.
(42) Bolton, P. H.; Kenner, R. D.; Khan, A. U. J. Chem. Phys. 1972,
57, 5604-5605.
For dibenzyl ketone and diphenyl disulfide ΦR ) 2Φesc, whereas
for the other radicals, ΦR ) Φesc
.

400 J. Am. Chem. Soc., Vol. 120, No. 2, 1998
Darmanyan et al.
Table 2. Calculated and Experimental Electronic Energy Gapsa
E, hartrees
∆E, cm-1
radical
ground stateb
excited statec
calcd
exptl11
% error
CH3O•
-114.70957
-437.35513
-131.90845
-418.18698
-150.50781
-795.81936
-473.17727
-189.68159
-302.72519
-628.51523
-114.55704
-437.21945
-131.88068
-418.10941
-150.47685
-795.78944
-473.1245
-189.65034
-302.70477
-628.41355
33559
29565
6956
17582
6683
6501
11548
6805
31614
26397
7580
18656
7030
7255
14367
7375
6
12
CH3S•
CH3CN+•
CH3CP+•
HOO•
-8
-6
-5
HSS•
-10
-20
-8
HSO•
CH3OO•
CH3COO2
•
4451
21822
5562
19328
-20
13
PhS• (2B2 f
2A2)
PhS• (2B2 f
2B1)
-628.51523
-628.50329
2581
PhOO•
-380.71867
-703.53534
-380.80040
-703.33882
6964
16903
PhSO•
17500d
a Calculated energies are RMP2/6-31G(d,p)//ROHF/6-31G(d,p). ROHF/6-31G(d,p) zero point energies, scaled by 0.9, are included. b Ground
state for all Cs species is 2A′′; ground state for C3V species is 2E. Convergence was achieved on the C3V species by removing symmetry constraints.
c Excited state for all Cs species is 2A′; excited state for C3V is 2A1. d Estimated from the onset of absorption from solution phase transient absorption.14
greater than those for PhS• and PhSO•. Strong experimental
evidence existed for low-lying excited electronic states for some
alkylperoxyl radicals, but the same was not true for PhS• and
PhSO•. Therefore, a computational approach was taken to
investigate the issue.
group in PhOO•, however, is directly conjugated. Nonetheless,
a quite similar excitation energy is calculated. The substituent
effect of the acetyl group is correctly predicted, so there is no
reason to believe the phenyl substituent effect is anything but
very small.
Because PhSO•, PhOO•, and PhS• are comparatively large
molecules, a relatively modest computational approach was
sought. A “test bank” of several radicals whose first electronic
excitation energies are known11 was used to help evaluate
methods: HOO• (Cs symmetry), HSO•(Cs), HSS• (Cs), CH3OO•
(Cs), CH3COO2• (Cs), CH3O• (C3V), CH3S• (C3V), CH3N•+ (C3V),
and CH3P•+ (C3V). For each of the Cs radicals, the ground state
PhSO• is considered next. Two absorption bands have been
observed by solution phase flash photolysis.14 The lower energy
band is broad and structureless, with a low extinction coefficient
and an onset at approximately 17 500 cm-1 (575 nm), far above
the singlet oxygen excitation level. However, this does not
answer the relevant question about the existence of a state lying
below 7880 cm-1, which was outside the range of wavelengths
investigated.
2
2
is A′′ and the first excited state is A′. For the C3V radicals,
2
2
the ground state is E and the first excited state is A1. None
of the radicals investigated has a low-lying quartet state.
Optimized geometries for each symmetry state were obtained
at the ROHF/6-31G(d,p) level, and single point energies were
found by using the RMP2 method (Table 2).43,44 The accuracy
of this method appeared to be sufficient for qualitative purposes
for the smaller radicals and PhS• (vide infra), with a root-mean-
square error of 12%. All of the Cs cases had energy gaps which
were underestimated, relative to the experimental values. Even
the worst case, HSO•, has an energy gap that is underestimated
by only 20%; a significant improvement over that is not found
until the CASSCF/cc-pVTZ level, at which point the calculated
energy gap is about 7% too low.45,46
The phenylsulfinyl radical is, like the other Cs species, found
to have a ground state of A′′. Its ground state is fully planar
2
and the unpaired spin resides in a π* orbital localized on the S
and O atoms.14 A limited configuration interaction calculation
(single excitations only) indicated that the lowest excited state
was composed almost entirely of a single configuration that
corresponded to the first 2A′ state, 19 500 cm-1 above the
ground state. The RMP2 energy difference between the lowest
2A′′ and 2A′ states for the relaxed planar conformations of PhSO•
is about 16 900 cm-1 (592 nm), consistent with the CIS
calculation and the absorption spectrum.
While the planar conformation of PhSO• is a minimum for
the 2A′′ state, a similar planar conformation is a transition state
The C3V energy gaps are underestimated less than those for
the Cs cases or are slightly overestimated. This may be an
2
for the A′ state. Rotation of the C-S bond by 90° gives a
2
new conformation of Cs symmetry. We shall refer to conforma-
artifact, as the E states do not converge in true C3V symmetry
2
tions of this nature as T-shaped. Along the A′′ surface, a
because of orbital degeneracy; therefore, they were allowed to
optimize without symmetry. As a result, the “2E” states have
an additional stabilization not available to the rigorously
symmetric C3V 2A1 states.
T-shaped conformation is the transition state for C-S bond
2
rotation. In contrast, along the A′ surface, the minimum is a
T-shaped conformation, and the planar conformation represents
the transition state for C-S bond rotation.
Because of this added complication, a more sophisticated
computational approach was taken for PhSO•. Using the
geometries obtained at the ROHF level for the A′′ and A′
states, CASSCF calculations were performed for both planar
and T-shaped geometries. Multireference second-order Møller-
Plesset (MRMP2) corrections were made to these energies
(Figure 4). These calculations gave similar gaps to the RMP2
energies shown in Table 2 and again lend credence to the rest
of the computational values it contains.
The data in Table 2 show the low-energy transition available
to the alkylperoxyl radicals. For systems such as PhCH2OO•
where a phenyl group is insulated from the peroxyl center by a
saturated carbon, little perturbation is expected. The phenyl
2
2
(43) Knowles, P. J.; Andrews, J. S.; Amos, R. D.; Handy, N. C.; Pople,
J. A. Chem. Phys. Lett. 1991, 186, 130-136.
(44) Lauderdale, W. J.; Stanton, J. F.; Gauss, J.; Watts, J. D.; Bartlett,
R. J. Chem. Phys. Lett. 1991, 187, 21-28.
(45) Xantheas, S. S.; Dunning, T. H., Jr. J. Phys. Chem. 1993, 97, 18-
19.
(46) Plummer, P. L. M. J. Chem. Phys. 1990, 92, 6627-6634.

Quenching of Singlet O2 by O- and S-Centered Radicals
J. Am. Chem. Soc., Vol. 120, No. 2, 1998 401
thoroughly.3,51-54 Aside from chemical reactions, four mech-
anisms are generally cited: energy transfer, charge transfer,
paramagnetically induced intersystem crossing, and electronic
to vibrational energy transfer (EVET). Charge-transfer quench-
ing occurs between electrophilic singlet oxygen and species such
as amines and sulfides that have a nucleophilic lone pair. By
and large, rate constants are e109 M-1 s-1, and many cases
are 10 to 1000 times slower than that. This is not an especially
likely mechanism for peroxyl radicals, since they are electro-
philic species.2,3,55
Paramagnetically induced intersystem crossing due to electron
exchange56 has been demonstrated for a series of nitroxide
radicals, whose kq values did not show any significant solvent
effect, but did have a large steric dependence.15 Rate constants
of 105 to 107 M-1 s-1 were observed. It is clear that, like EVET
for diamagnetic molecules, this mechanism will exist as a
“baseline” for all organic radicals. What remains is to ask if
other, faster quenching mechanisms also exist.
Energy transfer is, of course, most directly characterized by
observation of the excited quencher. Perhaps no case is more
prototypical or well-studied than â-carotene, whose kq is just
Figure 4. Energies of the planar and T-shaped PhSO• conformers in
2A′′ and 2A′ states, relative to the planar 2A′′ state. Calculations are at
the MRMP2/6-31G(d,p)//ROHF/6-31G(d,p) level.
over 1010 M-1 s-1 3,57,58
Other sorts of energy transfer quenchers
.
include square planar metal complexes of Ni(II) and Co(II) and
various cyanine-, azomethine-, and chlorophyll-type dyes. This
assignment is generally made on evidence of a sufficiently low-
lying triplet state, and rate constants in excess of about 1 × 109
M-1 s-1. In the following paragraphs, the suggestion is made
that peroxyl radicals represent a new general class of energy
transfer substrates for singlet oxygen.
The first two entries of Table 1 demonstrate that the rate
constants kq for quenching of singlet oxygen by PhS• and PhSO•
are no greater than about 108 M-1 s-1. Because the other
initiators (aside from 2 and 3) produce one of these two radicals
along with the carbon-centered radical, and because PhS• and
PhSO• have previously been shown to react slowly with ground
state oxygen (i.e., reaction 5), only the peroxyl radicals generated
in entries 3-11 contribute significantly to the quenching of
singlet oxygen, as indicated in the table.
The solution phase transient absorption spectrum of PhS•
consists of a sharp band at about 300 nm and a broad band in
the visible, but is somewhat solvent-dependent. Emission
corresponding to both absorption bands has been observed at
low temperature47,48 and in the gas phase.49 Data for gas phase
and low-temperature nonpolar matrixes are in reasonable
agreement, and the 0,0 band of the lowest energy emission is
at 19 330 cm-1
.
In C2V symmetry, there are four states that could be calculated
by the method used in Table 2. However, there is no guarantee
that these comprise the four lowest states. The ground state of
2
PhS• is B2 (defining the phenyl ring as the xy plane), and a
CIS run was performed without symmetry to get a handle on
the makeup on the low-lying excited states. Similar results were
obtained as found in the literature with a different basis set:50
the first three excited states were composed almost entirely of
2B1, 2A2, and 2B2 configurations, respectively. The 2A2 and 2B2
states are very similar in energy at this level of calculation,
21 000 and 22 600 cm-1 above ground state. Tripathi cites
Raman and LIF data to argue that the broad visible absorption
is due to more than one electronic transition, promotion to the
In the case of the chlorides 2 and 3 (entries 12-15), it is
necessary to take into account that photolysis leads to the
production of atomic chlorine, as well as the carbon-centered
radical.28,29,35,59 Chlorine atoms abstract hydrogen from aceto-
nitrile with a rate constant of 1.3 × 107 M-1 s-1 59,60
.
The
2
2A2 and the second B2 state.50 The method used to calculate
subsequent addition of cyanomethyl radical to molecular oxygen
(reaction 5) produces the cyanomethylperoxyl radical with a
values in Table 2 is only applicable to the former, but it is
reasonably close to the 0,0 band of the lowest emission.
rate constant61 of 1.3 × 109 M-1 s-1
.
2
Thus the photolysis of Ph3CCl in acetonitrile leads to the
formation two peroxyl radicals: Ph3COO• and NCCH2OO•. The
kq value for this initiator is 2-3 times higher than that for 6,
In the B2 ground state, the unpaired spin of the radical is
almost entirely localized on the p-orbital of the sulfur atom that
is in conjugation with the phenyl ring. Computationally, there
is a low-lying 2B1 state, although there is no experimental
evidence for it as of yet. In this state it is the p-orbital whose
axis is rotated 90° with respect to phenyl conjugation that is
singly occupied. Unsurprisingly, the calculated separation
(51) Lissi, E. A.; Encinas, M. V.; Lemp, E.; Rubio, M. A. Chem. ReV.
(Washington, D.C.) 1993, 93, 699-723.
(52) Gorman, A. A. AdV. Photochem. 1992, 17, 217-274.
(53) Bellus, D. AdV. Photochem. 1979, 11, 105-205.
(54) Gorman, A. A.; Rodgers, M. A. J. In Handbook of Organic
Photochemistry; Scaiano, J. C., Ed.; CRC Press: Boca Raton, FL, 1989;
Vol. II, pp 229-247.
(55) Neta, P.; Huie, R. E.; Ross, A. B. J. Phys. Chem. Ref. Data 1990,
19, 413-513.
(56) Hoyntink, G. J. Acc. Chem. Res. 1969, 2, 114-120.
(57) Farmilio, A.; Wilkinson, F. Photochem. Photobiol. 1973, 18, 447-
450.
2
2
between the B1 and B2 states is small, 2600 cm-1
.
Discussion
The quenching of singlet oxygen by various molecules has
been studied for a number of years, and has been reviewed
(47) Russell, P. G. J. Phys. Chem. 1975, 79, 1347-1352.
(48) Shibuya, K.; Nemoto, M.; Yanagibori, A.; Fukushima, M.; Obi, K.
Chem. Phys. 1988, 121, 237-244.
(58) Marston, G.; Truscott, T. G.; Wayne, R. P. J. Chem. Soc., Faraday
Trans. 1995, 91, 4059-4061.
(59) Bunce, N. J.; Ingold, K. U.; Landers, J. P.; Lusztyk, J.; Scaiano, J.
C. J. Am. Chem. Soc. 1985, 107, 5464-5472.
(49) Jinguji, M.; Imamura, T.; Obi, K.; Tanaka, I. Chem. Phys. Lett.
1984, 109, 31-34.
(60) Russell, G. A. J. Am. Chem. Soc. 1958, 80, 4997-5001.
(61) Neta, P.; Grodkowski, J.; Ross, A. B. J. Phys. Chem. Ref. Data
1996, 25, 709-1050.
(50) Tripathi, G. N. R.; Sun, Q.; Armstrong, D. A.; Chipman, D. M.;
Schuler, R. H. J. Phys. Chem. 1992, 96, 5344-5350.

402 J. Am. Chem. Soc., Vol. 120, No. 2, 1998
Darmanyan et al.
which produces only one kinetically important radical: Ph3-
COO•. From this, it is estimated that NCCH2OO• quenches
the fastest to the slowest. However, this is not as dramatic as
the steric effect observed for reaction 5, for which the rate
constant varies by an order of magnitude from Ph• to Ph3C•.14,37,38
Finally in support of the energy transfer hypothesis is some
of the evidence from the original report of quenching of O2-
(1∆g) by HOO• in the gas phase.8 The reaction vessel contained
O2(1∆g), O(3P), and a third component: ethylene, ammonia,
hydrazine, or hydrogen peroxide. Emission distinct from and
slightly to the red of that of O2(1∆g) was observed. In the
absence of any one of the three components, the new emission
was not seen, and it was deduced that only HOO• emission could
account for the observed spectrum. Since the new emission
was not observed without 1O2, it was concluded that the
fluorescent HOO• 2A′ state was achieved through energy
transfer.
Ordinarily, the connection between a gas-phase experiment
involving HOO• and solution phase experiments with radicals
such as PhOO• might seem tenuous. However, the contention
here is that it is the peroxyl functionality itself that is responsible
for the energy transfer, and that the substituent, be it H, Ph, or
Ph3C, is only of moderate consequence. Demonstration that
energy transfer occurs for HOO• should be sufficient to show
that energy transfer to the peroxyl radicals in Table 1 is
plausible.
singlet oxygen with a rate constant of ca. 5 × 109 M-1 s-1
.
Thus, with 3 as the initiator, a rate constant of approximately 2
× 109 M-1 s-1 can be estimated for the suberanylperoxyl radical
3-OO•.
The peroxyl radicals all quench singlet oxygen in solution
very efficiently, with rate constants kq of 2-7 × 109 M-1 s-1
.
By using the experimentally known 2A′′ f 2A′ transition
energies for HOO• and the small alkylperoxyl radicals as a guide,
it is quite likely that all of the peroxyl species have this low-
lying state. The nature of the transition is a promotion of an
electron from one of the doubly occupied distal O-atom lone
pairs to the half-filled O-O π* orbital. In other words, this is
a transition quite analogous to the carbonyl n f p* transition,
except that there is an extra electron in the system. A transition
in the region of 7000 cm-1 is thus inherent to the ROO•
chromophore itself. Even the conjugated PhOO• is calculated
to have a transition at an energy very similar to that of HOO•
and CH3OO•. Conjugation of the peroxyl to the carbonyl
function appears to lower the transition energy somewhat, but
there is no reason for the transition to disappear entirely or
become much higher energy with ordinary substituents. Be-
2
cause of the low-lying nπ* state and the comparatively high
Superoxide, O2-•, is related to HOO• by simple deprotonation,
and one might intuitively expect results for this species similar
to the peroxyl radicals discussed here. Reported experimental
values of kq for superoxide vary over orders of magnitude, with
rate constants for kq, we suggest that energy transfer is the major
quenching mechanism for alkylperoxyl radicals:
kq
9
• 2
-
• 2
O2(1∆g) + RO2 ( A′′)
8 O2(3Σg ) + RO2 ( A′) (8)
the highest being (7 ( 6) × 109 M-1 s-1 64-66
.
We do not have
any comment on the previous data, but it is important to point
out that a fundamental difference exists between superoxide and
all of the peroxyl radicals. Because it lacks a ligand to reduce
its D∞h symmetry to Cs, the frontier orbitals of a′ and a′′
symmetry that make up the transition in a peroxyl radical are
Reaction 8 is spin-allowed, and the rate constant of this
reaction could be close to the diffusion-controlled limit, kdif
)
3.7 × 1010 M-1 s-1 in acetonitrile.62 The reason for the
discrepancy is not clear, although even â-carotene does not quite
reach this limit. One possible explanation is that, for the peroxyl
radicals, energy transfer from singlet oxygen in acetonitrile is
so nearly thermoneutral that the observed rate is slightly below
the true energy transfer rate due to reversibility.63
2
degenerate in superoxide, and the ground state is Π. Experi-
mental and theoretical estimates are that the lowest energy
excited state for superoxide is a quartet some 2 eV (16 000
cm-1) above the ground state, clearly above the limit for energy
transfer from singlet oxygen.67 Charge transfer is thus probably
the major singlet oxygen quenching mechanism as previously
suggested.
The comparatively low upper limit on kq observed for
quenching by PhSO• inherently suggests that energy transfer is
not a likely mechanism and the calculations concur on this point.
The logical presumption is that singlet oxygen quenching occurs
by enhanced intersystem crossing, in analogy to the nitroxides.
Charge-transfer interactions cannot be ruled out.
Although the energy gaps for HOO• and CH3OO• are several
hundred cm-1 below that of singlet oxygen in the gas phase,
there is some reason to believe that the energy separation would
increase somewhat with solvent polarity. Chemical intuition
2
certainly suggests that the nπ* state would be less polar than
the ground state, and indeed calculated dipoles for both the
peroxyl and sulfinyl radicals are larger in the ground state. Use
of a solvent as polar as acetonitrile might induce a shift of as
much as 1000 cm-1 if precedent is to be taken from known
carbonyl photophysics. In Table 2, it can be seen that quenching
by PhCH2OO• is faster in hexane (dielectric constant ) 1.88,
viscosity ) 0.31 cP) than in CH3CN (dielectric constant ) 37.5,
viscosity ) 0.38 cP). Quenching is also apparently faster in
cyclohexane (dielectric constant ) 2.02, viscosity ) 1.0 cP)
than CH3CN despite the increased viscosity.
Like PhSO•, PhS• is a relatively inefficient singlet oxygen
quencher. Enhanced intersystem crossing is clearly an allowed
quenching mechanism, but the question remains as to whether
other mechanisms might be operative. The lowest energy
optically allowed transition (19 300 cm-1) is clearly too high
in energy to act as an energy acceptor from singlet oxygen.
Reaction rate constants for PhS• with various molecules show
that it is an electrophilic species.68 Hence electron-transfer
interactions with singlet oxygen are unlikely.
Consistent with this notion is that the kq values for acetyl-
•
•
peroxyl (CH3COO2 ) and benzoylperoxyl (PhCOO2 ) radicals,
determined by a related method, have been estimated at 1.2 and
1.5 × 1010 M-1 s-1 in benzene.12 As seen in Table 2, the gas-
phase transition energy for acetylperoxyl is about 2200 cm-1
below that of singlet oxygen, which will be low enough not to
be affected significantly by reversible energy transfer.
2
The calculations suggest an experimentally unobserved B2
level may very well reside only 2000 to 3000 cm-1 over the
ground state for PhS•. Nonetheless, rapid energy transfer
Steric interactions may also play a role in lowering the rate
constants. Sterics do have a modest effect on the rate constants
within the peroxyl radical series itself, a factor of about 3 from
(64) Mulazzani, Q. G.; D’Angelantonio, M.; Venturi, M.; Rodgers, M.
A. J. J. Phys. Chem. 1991, 95, 9605-9608.
(65) Guiraud, J.; Foote, C. S. J. Am. Chem. Soc. 1976, 98, 1984-1986.
(66) Rosenthal, I. Isr. J. Chem. 1975, 13, 86-89.
(62) Ware, W. R. J. Phys. Chem. 1962, 66, 455-458.
(63) Sandros, K. Acta Chem. Scand. 1964, 18, 2355-2374.
(67) Ewig, C. S.; Tellinghuisen, J. J. Chem. Phys. 1991, 95, 1097-1106.
(68) Ito, O. Res. Chem. Intermed. 1995, 21, 69-93.

Quenching of Singlet O2 by O- and S-Centered Radicals
J. Am. Chem. Soc., Vol. 120, No. 2, 1998 403
2
quenching of 1O2 is not observed, though it would be exothermic
by 4800-5800 cm-1. Energy transfer that has been slowed by
movement into the Marcus inverted region has been observed,69-71
but is very rare since the maximum rate constant in a series is
nearly always masked by the limitation of diffusion control.
Such an interpretation here might be attractive, but it is
extremely speculative.
peroxyl radicals are known to have A′ states approximately
7000 cm-1 above the ground state, which is very well matched
for energy transfer from singlet oxygen, about 7900 cm-1 above
its ground state. This excitation is an n f π* transition
analogous to carbonyl photochemistry. PhSO• has a similar
2nπ* excited state, but its energy is much higher than that for
the peroxyl case and is above that of singlet oxygen. Most
likely, the residual quenching of 1O2 occurs by the less efficient
paramagnetically enhanced intersystem crossing mechanism.
PhS• is also an inefficient quencher of singlet oxygen. No
optically detected state lies below 7900 cm-1, but calculations
suggest that there may be a very low-lying state, which is
apparently not an efficient energy sink for singlet oxygen.
Additionally, it is found that RMP2/6-31G(d,p)//ROHF/6-31G-
Finally, it is worth noting that the energy transfer interpreta-
1
tion for quenching of O2 by peroxyl radicals and the data in
Table 2 allows one to make a prediction not tested in this work.
Like the peroxyl radicals, HSS• has a very low-lying n f p*
transition at 7255 cm-1. It is therefore quite likely that it and
other perthiyl radicals would also be very rapid singlet oxygen
quenchers, unlike the other sulfur-centered radicals examined
here.
2
2
(d,p) calculations on A′′ and A′ states of several radicals
reasonably reproduce the experimentally determined energy
gaps.
Summary
Radicals have been generated in the presence of singlet
oxygen by the photolysis of aerated solutions containing both
various radical initiators and singlet oxygen sensitizers. For a
variety of peroxyl radicals, quenching rate constants of 2-7 ×
109 M-1 s-1 were determined. On the other hand, the
isoelectronic PhSO• radical is a much poorer quencher. Small
Acknowledgment. The authors at Iowa State are grateful
for support from the NSF (CHE 9708237) and Research
Corporation for this research.
Supporting Information Available: Optimized geometries
and absolute ROHF and RMP2 energies for the various radicals,
as well as MRMP2 and CASSCF energies for PhSO• (18 pages).
See any current masthead page for ordering and Internet access
instructions.
(69) Sigman, M. E.; Closs, G. L. J. Phys. Chem. 1991, 95, 5012-5017.
(70) Murtaza, Z.; Zipp, A. P.; Worl, L. A.; Graff, D.; Jones, W. E., Jr.;
Bates, W. D.; Meyer, T. J. J. Am. Chem. Soc. 1991, 113, 5113-5114.
(71) MacQueen, D. B.; Eyler, J. R.; Schanze, K. S. J. Am. Chem. Soc.
1992, 114, 1897-1898.
JA9730831